Bridge sensors that flag possible damage

How far away are bridges and tunnels that give early warning of failure?

Monitoring systems now run on some real bridges. They record strain, temperature and camera images, and can flag unusual behaviour. A 2023 case on a Swedish railway bridge shows a monitoring service flagging an abrupt change, after which an emergency inspection found a large crack in a truss member. Monitoring is still used case by case, and no source reviewed shows it routinely replacing or changing inspection across a bridge network. Sensors only see the points they are placed on, so inspection remains necessary.

AI illustration · Illustrative scene, not a photograph of the specific device or experiment.
WHERE WE AREPilots

Close. Early pilots are running in the real world.

Monitoring systems are installed and running on individual structures, and at least one documented case links an alert to an inspection that found a crack. The sources reviewed do not show monitoring as a routine part of managing a bridge network, and they describe monitoring on individual structures. That fits a pilot stage rather than scaling or widespread use.

The tech tree

Everything recorded on Vrulum that this future needs, oldest foundations at the top and the goal at the bottom. Lines show what each step requires.

What it takes: Bridge sensors that flag possible damage Unlocked Being researched Locked
  1. 2014Self-monitoring infrastructurePilot and early commercial Unlocked.
  2. 2018Distributed fibre-optic sensingPilot and early commercial Unlocked.
  1. 2008National long-term bridge performance programmeIn progress Being researched.
  2. 2008Foundation monitoring on a newly opened bridgeDone Unlocked.
  3. 2023Monitoring flags a change that leads to an inspectionDone Unlocked.
  4. Date not recordedReal-time monitoring project on a bridge in serviceDone Unlocked.
  5. Date not recordedRoutine use across a bridge networkNot started Locked.
  1. GoalBridge sensors that flag possible damageGoalRequires: Self-monitoring infrastructure, Distributed fibre-optic sensing, Real-time monitoring project on a bridge in service, Monitoring flags a change that leads to an inspection, Foundation monitoring on a newly opened bridge, National long-term bridge performance programme, Routine use across a bridge network. Locked.

The road so far

3 of 5 steps done, 1 in progress.

  1. DoneReal-time monitoring project on a bridge in service

    West Virginia's Structural Health Monitoring project put real-time monitoring on the East Huntington/Gunner Gatski Bridge, a cable-stayed bridge over the Ohio River carrying more than 14,000 vehicles a day. The project's implementation report, dated 25 June 2025, describes it as a SMART-funded effort whose objective is to use real-time data to help prioritise repairs and maintenance. [2]

  2. Done · 9 Mar 2023Monitoring flags a change that leads to an inspection

    On the Vänersborg Railway Bridge in Sweden, a cloud-based monitoring service detected an abrupt change in behaviour from the bridge opening on 9 March 2023 at 23:45. An emergency inspection found a large crack in a truss member, a partial crack affecting the bottom flange and part of the web. The authors describe the dataset as one of few actual registered damages for a structure in service, and they do not give the crack's length. [1]

  3. Done · 18 Sept 2008Foundation monitoring on a newly opened bridge

    On the I-35W bridge in Minneapolis, later called the St Anthony Falls Bridge, FHWA and partners instrumented foundations with strain gauges, temperature sensors and a field camera. Eight loaded dump trucks were driven in staged patterns across the bridge in September 2008 as part of truck-load testing. The bridge opened on 18 September 2008. The FHWA chapter notes that nearly one month of data were lost during a disconnection period, and that long-term live-load, diurnal and seasonal data were not yet available when it was written. [3]

  4. In progress · Apr 2008National long-term bridge performance programme

    FHWA's Long-Term Bridge Performance Program was launched in April 2008 as a planned 20-year effort, with periodic inspection, monitoring and NDE of a sample of bridges. The programme page is archived, was last modified on 5 July 2016, and says no update is needed. Whether the programme is still active is not confirmed by the sources reviewed. [4]

  5. Not startedRoutine use across a bridge network

    No source reviewed shows permanent monitoring being used across a bridge network to change inspection, maintenance or load decisions over a full service life. Owners would need to show that data continuity, sensor health checks, calibration and an agreed response process hold up over years. [4][2]

Why it matters

Inspection shows a bridge's condition only at the times and places it is carried out. Continuous data could, in principle, show problems sooner and point inspections to where they are needed. Poor sensor data or false alarms could also cause unnecessary closures or missed damage, so the value depends on whether the data change real decisions.

What stands in the way

  • The detection documented in the sources reviewed comes from one structure, so it does not show how such detections would perform across a network.
  • Sensor data need careful handling. A disconnection cost nearly a month of data on the I-35W bridge.
  • Interpreting a reading needs a model of the structure. The Vänersborg authors say a complete description of the structure was infeasible because the design is deficiently described in the available drawings.
  • Sensors measure only where they are placed. A reading from one point cannot show that the rest of a structure is sound, so inspection is still needed.
  • The national programme that would have tracked bridge performance over time is described as archived, so no current public national monitoring programme was confirmed in the sources reviewed.

Signals to watch

  • Published inspection results that confirm or contradict alerts from monitored bridges, including false alarms.
  • Long-term data from the West Virginia bridge, showing whether readings stay reliable after the initial project period.
  • Documented decisions, such as load limits, repair designs or replacement timing, that changed because of monitoring data.
  • Publicly reported lifecycle costs for sensors, communications, calibration and staff, compared with the cost of inspection alone.
  • Monitoring systems that report their own faults, such as failed channels, power loss or drift.

Sources

  1. peer reviewed paper
    Dataset for damage detection retrieved from a monitored bridge pre and post verified damage
    John Leander, Jacob Nyman, Raid Karoumi et al.. Data in Brief, 26 Oct 2023.
  2. government report
    West Virginia Structural Health Monitoring Project Implementation Report
    West Virginia Department of Transportation, Division of Highways (SMART Program, via ROSA P), 25 Jun 2025.
  3. government report
    Chapter 4: I-35W Bridge Foundation Monitoring, in State of the Practice and Art for Structural Health Monitoring of Bridge Substructures (FHWA-HRT-09-040)
    Federal Highway Administration, May 2014.
  4. government report
    About the Long-Term Bridge Performance Program
    Federal Highway Administration, Jul 2016.

Last reviewed 21 Oct 2026.